Stereo vision involves the extraction of three-dimensional information from images by comparing information about a scene from two different viewpoints. Traditional stereo algorithms are used to generate depth maps from color images.